Tech Hiring 2026 Runs Through Banks, Factories and the AI Labs

In March, OpenAI told its staff it plans to grow from roughly 4,500 people to around 8,000 by the end of 2026, with most of the new seats in engineering, research, product and enterprise sales.

You won't see that in a layoff tracker, and if the trackers are where you get your picture of tech hiring in 2026, you'd reasonably conclude nobody is hiring at all. The data disagrees. Software postings have risen for fifteen months. Robert Half found 78% of technology leaders planning to add permanent headcount in the second half of the year. LinkedIn says the fastest-growing job title in the United States is AI engineer.

We don't want to oversell it. The demand is selective and leans hard toward experienced specialists, and generalist pay has come down from the 2022 peak. But selective demand still pays the rent, and it goes to people who know where to look.

The AI labs are adding people as fast as big tech cuts

OpenAI isn't alone. Anthropic's open postings roughly doubled year over year, and its workforce has grown more than tenfold since 2023. These are the companies building the technology everyone else is restructuring around, and they're hiring while the giants trim.

Two things about that hiring should interest you even if you'd never apply to a lab. A big share of it is customer-facing engineering: forward-deployed and solutions roles whose job is to make models work inside real businesses. Strong generalists with domain knowledge can get those jobs, and you don't need a research background.

The other thing is price. Lab hiring sets what AI skills are worth across the whole industry, and it's one reason PwC measured the AI wage premium widening to 56%. When the labs bid up a skill, the bank down the street has to pay more for it too.

Banks and manufacturers are doing much of the hiring

Robert Half's 2026 survey found technology hiring strongest in financial services and in manufacturing and distribution, rather than at technology companies. Banks and insurers are modernizing old platforms and buying AI capacity at scale. Manufacturers want automation, supply-chain software and industrial data, and they're often building their first in-house engineering teams to get it.

Healthcare, defense, aerospace and government have something useful in common, which is that their demand doesn't swing with interest rates or with a CEO's efficiency memo. Our calculator has scored those industries as sheltered since 2024, and this year's hiring data backs that up. They're also where AI-governance rules are adding roles instead of removing them.

August's jobs report shows the split neatly. The US economy added 162,000 jobs and unemployment held at 4.1%, while the information sector shed jobs the same month. The economy around tech is in decent shape, and it's buying tech skills.

The job titles are shifting as well. LinkedIn's Jobs on the Rise list for 2026 put AI engineer first in the US, and AI, machine-learning and data-science postings more than doubled between 2024 and 2025. Indeed now finds AI mentioned in 6.3% of all US job ads, nearly twice the 2022 peak.

The more interesting growth sits one step over from those headline titles. Employers told Robert Half that the capability they find hardest to hire is 'AI adoption and automation workflows', meaning someone who can decide which processes AI should absorb, redesign them, measure what happened and retrain the team. On an org chart, that's an AI product manager or an AI enablement lead. People from operations, product, analytics and management backgrounds can do that job. Add evals engineers, AI-governance specialists and forward-deployed engineers, and you have a set of titles that weren't on job boards three years ago.

Specialist pay has held and generalist pay has reset

Robert Half's 2026 starting ranges run $134,000 to $193,250 for AI/ML engineers, $127,000 to $180,750 for data engineers and $118,500 to $190,750 for cybersecurity engineers, with software engineers at $109,250 to $175,500. Kafka, Databricks and Azure show up in the same surveys among the scarcest skills, which in practice means they carry a premium.

Generalists get the less pleasant news. Several analyses of 2026 offers put generalist engineers' base salaries 15% to 25% below their 2022 peaks, pushed down by AI tooling, contract work and a big pool of recently laid-off people. We'd treat those analyses as a range rather than a precise figure, but the direction is consistent. A generalist with one specialty they can prove gets paid like a specialist. A generalist without one ends up competing on price.

So if you're aiming at the growing part of the market, search by sector rather than by logo. Pick one specialty with a scarcity premium, whether that's data engineering, platform, security, AI engineering or AI product, and build one public piece of proof in the next ninety days. If your current job lets you redesign a workflow around AI, do it and record the hours saved, since that's the capability employers keep saying they can't find. One of the named tools, Kafka or Databricks or Azure, will likely do more for your response rate than a fifth general certification.

Before you commit to any of it, run the pair through our role comparison tool to check the demand gap, and look at the Opportunities page for the current, sourced list of who's hiring.